Associations between DXA-derived parameters of body composition and metabolic dysfunction-associated steatotic liver disease in patients with T2DM and prediabetic states.

Background: Sarcopenic obesity is closely related to metabolic dysfunction-associated steatotic liver disease (MASLD), but the independent contributions of lean mass and fat mass components to MASLD are not well understood. Our study aimed to evaluate the relationship between the dual-energy X-ray absorptiometry (DXA)-derived soft tissue components and the extent of liver steatosis in patients with MASLD.

Methods: A cross-sectional study of 118 obese/overweight patients aged 33-78 years, with type 2 diabetes mellitus (T2DM) or prediabetes and MASLD, on oral antidiabetic medication was conducted. Sex-stratified correlation analysis was performed between DXA-derived lean mass and fat mass parameters, (e.g., relative muscle mass [RMM], lean mass/ fat mass [LM/FM] and appendicular lean mass [ALM]), as well as between each of those parameters and the hepatic steatosis index (HSI). Multiple linear regression models were fitted with android fat percentage as the dependent variable, and lean mass indices as independent variables. The models were adjusted for age, sex, the HOMA index, triglyceride and ALT levels. Accordingly, ROC curves were plotted with HSI=36 as a classifier of steatosis.

Results: A significant negative correlation was detected between android fat % and RMM (r=-0,96, P≤0.001), between android fat percentage and ALM /BMI (r=-0.70, P≤0.001), between android fat percentage and ALM /height (r=-0.28, P≤0.001); between HSI and RMM (r=-0.50, P≤0.001); and between HSI and ALM/BMI (r=-0.39, P≤0.001). The significant positive correlations were as follows: android fat percentage and BMI (r=0.53, P≤0.001); android fat % and ALT (r=0.25, P=0.04); HSI and android fat percentage (r=0.59, P≤0.001); HSI and gynoid fat % (r=0.39, P≤0.001). The AUCs for android fat percentage in the models calculated from LM/FM were as follows: adjusted for the HOMA index, age and sex group, ROC=0.748 (95% CI 0.66-0.83); adjusted for ALT and sex group, ROC=0.743 (95% CI 0.66-0.83). The AUC for android fat percentage in the model calculated from RMM: adjusted for triglycerides, ALT and sex group, ROC=0.741 (95% CI 0.65-0.83).

Conclusions: Our findings demonstrate that an increase in android fat distribution and a decrease in lean mass are positively associated with MASLD. The regression models support the utility of DXA-derived indices as practical, indirect markers of liver steatosis for clinical application in patients with TDM2 and MASLD.